Information Technology Reference
In-Depth Information
networks share remarkable percentage of overlapped users.
1
The overlapped users
serve as both index for organization, and bridge for association between heteroge-
multimedia data are essentially “user-centric”, which are generated from user contri-
bution and analyzed for customized user services. The contributed overlapped user is
a natural and efficient index for organization, which will then facilitate personalized
user services. On the other hand, the overlapped users' interactions with heteroge-
neous social multimedia data provide important clues for association mining. Instead
of analysis from the scratches, the interactions can be viewed as high-level supervi-
sion, where the wisdom of crowds are exploited for association mining.
The three basic tasks of user-centric social multimedia computing, can all be
extended under the cross-network scenario. In this chapter, corresponding to the
above challenge in organization and association, we elaborate cross-network social
multimedia computing, by introducing two illustrative work on user modeling and
multimedia knowledge mining, respectively.
5.2 Related Work
5.2.1 Macro Cross-Network Analysis
With various social media networks growing in prominence, analyzing and exploiting
the characteristics and correlations between social media networks have attracted
attentions recently. Among the related work, we categorize the ones in network-
centric view and having no consideration into the explicit correspondence between
the accounts to specific users as macro cross-network analysis.
One research line in macro analysis is to examine the characteristics of different
social media networks. For structural characteristics, in [
19
], degree distribution,
clustering coefficiency, and evolution over time are investigated to validate network
properties, such as power-law, small-world, and scale-free, in different social media
networks. In [
18
], traditional Social Network Analysis (SNA) measures, such as
degree centrality, shortest path, are extended under the cross-network circumstances.
For user activity patterns in macro-level, Guo et al. have investigated how users par-
ticipate in and contribute to blogging, social bookmarking and question answering
social media networks, respectively [
12
]. With different user activity patterns dis-
covered in the three types of networks, this work lays out an analytical foundation
for further cross-network user activity analysis. In [
7
], the authors studied the moti-
vational factors of users participating into various social media conversations, with
observations that in different social media networks, users enjoy different mixtures
of intrinsic and extrinsic motivational factors.
1
Please refer details to “Anderson Analytics 2009 report: what your favorite social network says
about you?”.
